Statistical power simulations of the gene expression dataset.

The statistical power (Z; 1- ß) is the probability of obtaining statistical significance in comparing gene expression. Simulations for unpaired two-class comparison statistics are described in the Materials & Methods. Calculation includes the following parameters: - a significance level (α) of 0.01. - the observed biological variability (σ). - a sample size (n) from individual [PIP++], [PIP+] and [PIP−] samples. - a true difference (i.e. 1.2, 1.35, 1.5 or 1.8) in mean expression ratios between the respective classes.